OPTICAL ANTI-ALIASING FILTERS
Optical Filtering for Aliasing, Moiré, and Modern Small-Pixel Sensors
Mosaic Optoelectronics develops optical anti-aliasing filters for imaging systems where sensor sampling can produce moiré, false detail, color artifacts, and other spatial errors. Our birefringent designs condition image-forming light before it reaches the sensor through controlled optical redistribution rather than indiscriminate blur.
THE IMAGING PROBLEM
When Fine Scene Detail Interacts With the Sensor
Digital sensors sample an optical image at discrete pixel locations. When scene detail approaches or exceeds the spatial frequencies the sensor can sample correctly, artifacts may appear that are not present in the original scene.
SMALL-PIXEL SENSORS
Smaller Pixels Raise the Sampling Limit — They Do Not Eliminate Aliasing
As pixel pitch decreases, the sensor can sample finer spatial detail. That increases the Nyquist frequency and can reduce aliasing from structures that would have exceeded the sampling capability of a larger-pixel sensor.
But aliasing does not disappear simply because pixels become smaller. If the lens, magnification, wavelength, and scene still deliver useful image information above the sensor's sampling limit, those frequencies can still fold back into the captured image as false structure, moiré, or color artifacts.
For modern high-resolution cameras, the question therefore becomes less whether the sensor has small pixels and more whether the complete optical system is delivering spatial information the sensor cannot represent correctly.
SYSTEM-LEVEL DESIGN
Pixel Pitch Is Only One Part of the Aliasing Problem
The sensor does not determine aliasing behavior by itself. The image presented to the sensor is shaped by the lens, magnification, aperture, wavelength, object detail, focus, and other optical conditions.
Sensor
Pixel pitch and sensor geometry determine the spatial sampling grid and establish the sensor's Nyquist limit.
Lens MTF
The lens determines how much high-spatial-frequency contrast actually reaches the image plane.
Scene Detail
Repeating traces, grids, fibers, edges, textures, and fine patterned structures can contain strong spatial-frequency content.
Magnification & Wavelength
Object-space dimensions are transformed into image-space detail, while wavelength and aperture influence optical resolution.
OPTICAL LOW-PASS FILTERING
Managing Spatial Information Before Sensor Sampling
An optical anti-aliasing filter modifies the image before it reaches the sensor so that spatial content most likely to produce sampling artifacts is reduced or redistributed.
In birefringent filter designs, controlled beam displacement can distribute image information over neighboring sensor locations. The amount and direction of displacement depend on optical material, crystal orientation, thickness, wavelength, polarization behavior, and stack configuration.
The objective is not simply to make the image softer. The filter must provide enough spatial conditioning to reduce objectionable aliasing while preserving useful in-band image information.
STRUCTURED OPTICAL CONDITIONING
From Generic Blur to Controlled Spatial Redistribution
The geometry of the point-spread response can be engineered around the sampling problem rather than treated as uncontrolled image softening.
A birefringent OLPF can be understood as a device that deliberately structures how image-forming light is distributed before it reaches the sampling grid.
Each birefringent element can create a controlled displacement between optical components. By arranging multiple elements with selected orientations and thicknesses, the resulting image energy can be distributed into an engineered spatial pattern rather than simply spread in an uncontrolled way.
This gives the filter designer another degree of freedom: not just how much the image is spread, but where that energy is placed relative to neighboring sensor locations.
In this sense, the filter acts as a form of pre-sensor optical conditioning, shaping the point-spread behavior of the imaging system before discrete sampling occurs.
BIREFRINGENT OPTICS
Controlled Beam Displacement
Birefringent materials can separate light into polarization components that experience different optical behavior inside the crystal.
In an anti-aliasing filter, this behavior can be engineered to produce controlled lateral displacement of image information. Multiple birefringent elements may be arranged to create a desired spatial distribution in more than one direction.
Crystal orientation, material properties, element thickness, wavelength, incidence angle, and polarization state all influence the resulting displacement and therefore the filter's point-spread behavior.
MULTI-ELEMENT ARCHITECTURES
Building an Engineered Sampling Pattern
Successive birefringent elements can progressively build a more complex spatial distribution before a final polarization-control element sets the emerging polarization state.
More than one birefringent element can be combined to create a controlled two-dimensional distribution of image energy.
Successive elements can introduce additional displacement directions or separations, allowing a designer to shape the effective point-spread function presented to the sensor.
Depending on the application, the stack may also incorporate polarization-control elements such as retarders to manage the polarization state emerging from the filter.
The resulting architecture can therefore combine spatial redistribution and polarization management within a compact optical stack.
MODERN SENSOR DESIGN
Higher Sensor Resolution Makes Optical-System Matching More Important
Small pixels allow an imaging system to capture finer detail, but realizing that capability requires the optical system and sampling system to be considered together.
More Spatial Bandwidth
Smaller pixels raise the sensor sampling frequency, allowing finer image structure to be represented before aliasing begins.
Better Lenses Matter More
High-performance optics may continue transmitting meaningful contrast near or beyond the sensor's Nyquist frequency, particularly in demanding machine-vision and inspection systems.
Condition Only What Is Necessary
A properly matched OLPF can reduce problematic out-of-band content while preserving as much useful spatial information as practical.
FILTER DESIGN
The Correct Filter Depends on the Complete Imaging System
Optical anti-aliasing behavior cannot be selected from sensor resolution alone. The useful design depends on the interaction between the sensor, lens, scene content, wavelength, polarization requirements, and required image performance.
Sensor Sampling
Pixel pitch, sensor geometry, color-filter architecture, and sampling characteristics influence where aliasing may become significant.
Optical System
Lens MTF, magnification, aperture, wavelength, focus, and object detail determine the spatial information delivered to the image plane.
Required Filter Response
The desired balance between retained image detail and artifact suppression determines the useful spatial redistribution or point-spread function.
SYSTEM INTEGRATION
Designed Around the Available Optical Path
Anti-aliasing filters may need to fit within an existing lens, camera, sensor package, or custom optical assembly. Mechanical integration is therefore part of the optical design problem.
APPLICATION-SPECIFIC CONFIGURATIONS
Custom Filter Geometry and Optical Behavior
Many anti-aliasing requirements do not map cleanly to a single standard filter configuration.
A filter may involve multiple birefringent elements, polarization control, wavelength-specific considerations, mechanical packaging, or other optical functions depending on the imaging system.
For custom work, the starting point is the imaging requirement rather than a fixed catalog configuration.
APPLICATIONS
Where Optical Anti-Aliasing Can Be Useful
Optical anti-aliasing is most relevant where fine repeating detail, high-performance optics, or high-spatial-frequency scene information can interfere with reliable sensor sampling.
Machine Vision
Industrial imaging systems evaluating edges, textures, repetitive features, fine grids, or patterned structures.
Electronics Inspection
Imaging of circuit patterns, component arrays, conductors, traces, solder features, grids, and other repetitive structures.
Scientific & Metrology Imaging
Imaging systems where sampling artifacts could interfere with measurement, interpretation, subpixel analysis, or quantitative visualization.
Small-Pixel High-Resolution Cameras
Specialized cameras where sensor sampling, lens MTF, and fine scene detail must be evaluated together as part of the optical design.
High-resolution electronics or semiconductor inspection image showing dense repetitive structures where aliasing risk is easy to understand visually.
NON-BLURRING COLOR ANTI-ALIASING
A Different Approach to Color-Alias Suppression
Spatial low-pass filters intentionally shape high-spatial-frequency content. Mosaic has also developed proprietary methods aimed specifically at color-aliasing artifacts in Bayer-pattern and similar color-filter-array sensors.
Sensor-Aware
The approach is more tightly coupled to the underlying sensor architecture than a generic spatial low-pass filter.
Different Principle
It operates by principles different from conventional spatial-frequency bandlimiting rather than simply increasing optical blur.
Sharpness Preserved
The objective is to suppress color-aliasing artifacts without the conventional loss of image sharpness associated with stronger spatial low-pass filtering.
CUSTOM POINT-SPREAD FUNCTIONS
Beyond Simple Four-Spot Architectures
Complex and application-specific point-spread functions can be developed to shape how image-forming light is distributed before sensor sampling.
For Cinema and Broadcast
More complex spot patterns can be used to address secondary and more complicated color-aliasing modes while preserving useful image detail.
For Machine Vision and Metrology
Application-specific spatial response can strengthen suppression of problematic frequencies while preserving true in-band information needed for quantitative imaging and subpixel measurement.
Structured Image-Plane Conditioning
These patterns illustrate the broader design principle: the filter can deliberately redistribute image energy into a selected spatial arrangement rather than treating anti-aliasing as generic blur.
MATERIAL OPTIONS
High-Birefringence Materials for Unusual Geometries
When application requirements justify them, filter designs can use high-birefringence substrate materials such as lithium niobate or KTP.
These materials can enable larger beam separations or thinner optical structures where available optical-path space or aberration control makes conventional geometries impractical.
Photograph of lithium niobate, KTP, or another high-birefringence optical component with a useful scale reference.
STARTING A FILTER PROJECT
Useful Information for Evaluating an OLPF Requirement
The more information available about the imaging system, the easier it is to determine whether optical anti-aliasing is appropriate and what spatial response may be useful.
For small-pixel and high-resolution systems, sensor pixel pitch alone is not enough. Lens performance, magnification, wavelength, object detail, and representative images are especially valuable.
- Sensor model, format, and pixel pitch
- Image examples showing the aliasing or moiré problem
- Lens model or available MTF information
- Magnification and working distance
- Operating wavelength or spectral range
- Aperture or f-number where relevant
- Typical object features or repeating structures
- Available filter dimensions and optical-path space
- Polarization requirements, if applicable
- Desired balance between image detail and artifact suppression
- Prototype and expected quantity requirements
OPTICAL ANTI-ALIASING
Designing Around a Small-Pixel Sensor or Seeing Sampling Artifacts?
Share the sensor, lens, wavelength, optical configuration, and representative image artifacts. We can help evaluate whether structured optical conditioning or another anti-aliasing approach is appropriate for the application.